Comments/Ratings for a Single Item

If an email gets returned to you as undeliverable, you can go mark the address it was sent to as dead in the Person table of the database, which will cause the display of the email address to be replaced with a notice that it is not working and should be changed. It's the last column in a row, and you can change its value to 1.

One thing that will need to be addressed before the future rounds is Berolina Chess.  The rule enforcing preset does not work so someone will have to fix it or we will have to use a non rule enforcing one.  I can probably do it, but with each preset, I'm spending quite a bit of time learning how to do things and I'm no longer on vacation, so I don't have the few hours a day to research and learn.  I believe all of the other presets are good - I did not check Wildebeeste, though.  I'm not sure if it is a rule enforcing one or not.

Sure, SxG, you're welcome to join.

I have confirmation from everyone except:

Vitya Makov
John Davis
Daniel McDuff

I am still waiting for confirmation from:

Vitya Makov
John Davis
Daniel McDuff

Please let me know in the next day or so.  When I make the pairings anyone who hasn't confirmed won't be included.

Thanks,
Greg

Confirm.

While I am sitting this tournament out, I wish to ask Fergus, for the sake of the tournament, alone, whether a bug in the Capablanca Chess preset (that I noted in a Kibbitz Comment to an old game I had with Carlos) has been fixed or not.

Kevin

Refresh my memory, Kevin. I don't remember what bug you mentioned.

Since Antoine is not around to update the Berolina Chess preset, I have started on a new one, which will display legal moves properly. I'll see if I can get it done today.

Hi Fergus

The Capablanca Chess log for my game with Carlos was this one:

http://play.chessvariants.com/pbm/play.php?game=Capablanca%20Chess&log=panther-cvgameroom-2017-14-164

If you play my last king move of the game with 'View' you'll see Carlos' king is dragged one cell in the same direction as a result. At that point, after many days had passed, I offered him a draw (by person to person regular email) so he wouldn't lose on time if he didn't try any move. When you view the last move 'Drawn' you'll then see that Carlos' king is somehow put back on the proper cell as a result.

P.S.: Tournament participants have no need to worry, I think, as I now recall Capablanca Chess was not one of the final choices made for variants to be played in this tournament. My apologies for any confusion.

Kevin

unfortunately Daniel hasn't confirmed leaving us with an odd number.  It occurs to me now that I try to make the pairings that my scheme doesn't really work with an odd number.  I could include byes but 2 people wouldn't get a bye so that's not really fair.

Don't suppose anyone else wants in?

Kevin, whether the bug in Capablanca's Chess could affect the tournament depends on where it is. Although it doesn't share any code with Victorian Chess, which is the most similar game, it does use the chess include file, which might be used by other games in the tournament.

Greg, you can handle any number like this. List all participants, copy the list and rotate it by one person. Copy and rotate it again for each additional game. So long as the games and players are equal in number, pairing the original list with a rotated list will give you unique pairs of players for each game.

The bug discovered in a game of Capablanca Chess is now fixed. It was in castlepos, which was in the chess include file. It made a potential King move, verified that it was safe, and then restored the position. But verify [condition] is equivalent to "if [condition] is false, return false". So, if the King was not safe, and the Black King was not at g8 thanks to the White Chancellor on the exposed g file, it would exit the castlepos subroutine before it had restored the position, which resulted here in an erroneous move by the Black King to an unsafe position.

Unfortunately, we have 9 games and 11 participants.

It would work with 9, 10, or 12 participants, but it's not clear to me how to make it work with 11 (without adding 2 more games)

The same bug that was in the chess include file had also been copied into the chess2 and chess3 include files, which I have now fixed. Since most of the programmed presets use one of these include files, this bug was widespread among several games. Thanks to spotting it one game, it is now fixed for several games.

So long as there are more participants than games, this method will create unique pairings, so that no one ever plays anyone else more than once, and so long as the difference between players and games isn't too great, it will assure that players who don't play against each other will still play against several of the same opponents. If you hope for everyone to play everyone, then you will need to add more games. But if you are content with everyone playing only nine games, then this method will allow you to distribute players for every game, no matter whether there is an odd or an even number of participants.

Scrap what I said. This would work only if everyone plays the same game twice, which would double the number of games played.

I think fixing that bug messed up my preset of Anarexic Chess.  Can't make any moves now.

I have finished writing a new preset for Berolina Chess. This one uses chess3, and like chess3, it displays legal moves and also enforces three-times repetition and the 50-moves rules.

I don't see any preset for anything called Anarexic Chess. Give me a link if you want me to look at it.

Sorry, guys, but I've hit another snag.

The odd number of players doesn't work, so I was going to drop SxG, on the basis that he was the latest joiner, so he should be the one to be dropped.

Then I turned to generating the pairings.  Been struggling with it for hours, with no success.  Eventually I discovered that the problem was not solvable.  The parameters: 10 players, 9 games, no players play each other more than once.  Seems like this is possible, but it's not.  Number of games = 3 rounds of 3 games with 5 pairings in each = 45 total games.  What are the number of unique pairings?  9+8+7+6+5+4+3+2+1 = 45!  Ok, so 45 games and 45 unique pairings - great, right?  No, because there is the additional restriction that the pairings are grouped into the 9 games.  Not possible.

This problem is surmountable if we allow pairs of players to play more than one game.  But it's late, and I'm pretty intoxicated, and the mathematics elude me at the moment.  So I'll attack this tomorrow.  Maybe, if we're really lucky, Daniel will turn back up by then and confirm his participation, and the whole problem will go away, as with 12 players, the number of pairings will far exceed the number of games.

Incidentally, if anyone knows how to contact Daniel MacDuff, that would be awesome.

Hey Greg!

It could be not a matter of forcing the tournament into certain parameters, but maybe designing the parameters around the tournament.  Without taking too deep of a look into it (as it is late here too and I wish I was intoxicated), maybe adding 2 more variants to the mix.  With 11 games, everyone plays everyone once and gets one bye.  (4 rounds - 3 games/each and 2 in the last or 3 rounds - 4 games/each and 3 in the last)

Also, subtracting variants to 6 and doubling the games played (one black and one white) in each variant should yield the same amount of games played with 1 bye each if my math is correct.  (3 rounds - 2 variants/round  and 2 games/variant)

Alternatively, we can just start the tournament with Daniel in the games, as he committed previously.  He could be on a short vacation or have wrong contact info or family emergency.  I'd be upset to join a tournament and then find out that I wasn't in it, in spite of attempts to contact me.  

I know that all of this kind of trounces the intended spirit of the tournament, but these are just a few ideas that could end the suffering for you.  I'm down for whatever, but one thing that I don't feel is right is to cut someone from the tournament after giving them the green light to participate.  

Thanks, Nick, for the ideas.  Just running with Daniel included is certainly one possibiliy, even though he's disappeared.  If he didn't re-appear, it would wind up being a bye for almost everyone, but not everyone, so not completely fair.

Adding two more games and everyone gets a bye doesn't work.  I think we have the same pairing problem.  We'd essentially have 12 players, because you'd consider "bye" a player.  We'd have 66 games (counting a game against "bye" as a game.)  And we'd have 66 possible pairings.

Ok, the pairings worked cleanly with 12 players and the existing 9 games, so that's what we're going with.  SxG and Daniel MacDuff are both in.

I'll update the web page to show the final list of games, players, and pairings.  (I still need to work on balancing it so each player has the white pieces in either 4 or 5 games.)  Then I'll create the actual games in GC.

Thanks, everyone, for bearing with me.  I didn't think making the actual assignments would be difficult.  Clearly I need to crack open my "Discrete Mathematics with Applications" textbook which fortunately I saved.